unit selection in chart does not respect advanced function calculation
Hey,
I created a report with three columns:
column 1, the date;
column 2, a raw value over time;
and column 3 (from a subquery) which used advanced function ('percentage against column 2') to get a percentage value over time.
I wanted to use the unit picker in a combination chart, so a user could look at daily/weekly volumes and % etc.
However, I noticed that the date picker simply aggregates the values output by the initial query, and loses context of where the calculated value comes from. So, things like percentage against column come out wrong, as the aggregation does not take into account the weighting each value should have, when aggregating.
As an example, if I was aggregating over two days
date count %
01/04 100 20%
02/04 10 40%
then looking at them combined I would expect count: 110 (this is ok) and %: ~22%
BUT I get %: 30% (if aggregation is set to average in charts) on the chart.
So I guess the easiest thing to do here is to prevent people from being able to enable the unit picker when using advanced functions, as this can lead to incorrect data being displayed.
Hope this makes sense!
Thanks,
Jonas
The same problem 
Hi Jonas,
Thank you for reaching out. I am able to replicate what you're seeing and I do think it is misleading and am in agreement that we either need to make it work correctly or don't allow people to do it!
I've logged this as a defect in our internal tracking system and will post any potential updates regarding this here.
Regards,
Mike
Hi Jonas,
Thank you for reaching out. I am able to replicate what you're seeing and I do think it is misleading and am in agreement that we either need to make it work correctly or don't allow people to do it!
I've logged this as a defect in our internal tracking system and will post any potential updates regarding this here.
Regards,
Mike
HI, is there any update about this issue?
HI, is there any update about this issue?
What a useful article! You provide me with a lot of crucial information. I greatly appreciate it.
wordle
What a useful article! You provide me with a lot of crucial information. I greatly appreciate it.
wordle
It sounds like the unit picker in the combination chart is not taking into account the weighting of the calculated values when aggregating. This is causing the percentage against column to come out wrong, especially when aggregating cookie clicker over multiple days.
It sounds like the unit picker in the combination chart is not taking into account the weighting of the calculated values when aggregating. This is causing the percentage against column to come out wrong, especially when aggregating cookie clicker over multiple days.
It's crucial to ensure accurate data aggregation to avoid confusion, much like in the Bad Parenting game where missteps lead to unintended consequences. Preventing unit picker access during advanced functions could simplify data analysis and improve clarity. By maintaining a consistent approach, we can enhance overall user experience and ensure reliable data management. What are your thoughts on other potential solutions to this issue?
It's crucial to ensure accurate data aggregation to avoid confusion, much like in the Bad Parenting game where missteps lead to unintended consequences. Preventing unit picker access during advanced functions could simplify data analysis and improve clarity. By maintaining a consistent approach, we can enhance overall user experience and ensure reliable data management. What are your thoughts on other potential solutions to this issue?
You must maintain your wits about you in Slope Game. Your speed and the difficulty of avoiding obstacles both grow as you roll down the slope.
You must maintain your wits about you in Slope Game. Your speed and the difficulty of avoiding obstacles both grow as you roll down the slope.
Hey everyone, I've noticed a peculiar bug. When using advanced functions in charts, the unit selection seems to ignore the calculated results. It's almost like the chart is stuck in its own little Fnaf jumpscare, frozen in place, not reflecting the actual data derived from the function. Anyone else experiencing this inconsistency? It makes accurate data analysis quite a challenge.
Hey everyone, I've noticed a peculiar bug. When using advanced functions in charts, the unit selection seems to ignore the calculated results. It's almost like the chart is stuck in its own little Fnaf jumpscare, frozen in place, not reflecting the actual data derived from the function. Anyone else experiencing this inconsistency? It makes accurate data analysis quite a challenge.
Hey Jonas,
Thanks for sharing your experience with the unit picker issue! It’s frustrating when calculated values lose context, especially in a visualization setting. Your suggestion to restrict the unit picker when advanced functions are involved seems like a solid approach to prevent confusion. By the way, if you're into engaging visualizations, check out my favorite snow sports simulation, Snow Rider 3D – it’s a fun way to combine gaming with data. Keep up the good work!
Cheers!
Hey Jonas,
Thanks for sharing your experience with the unit picker issue! It’s frustrating when calculated values lose context, especially in a visualization setting. Your suggestion to restrict the unit picker when advanced functions are involved seems like a solid approach to prevent confusion. By the way, if you're into engaging visualizations, check out my favorite snow sports simulation, Snow Rider 3D – it’s a fun way to combine gaming with data. Keep up the good work!
Cheers!
Yeah, that makes perfect sense, and I’ve run into similar issues with percentage-based metrics before. The unit picker is great for raw counts, but once you start mixing in calculated fields—especially ones that depend on weighted relationships—it just averages everything blindly and the logic golf hit behind the percentage gets lost. Your example highlights exactly why the results look off. Disabling the unit picker when advanced functions are used sounds like a reasonable safeguard, or at least there should be a warning so users know the aggregated percentages won’t be meaningful. Good call bringing this up.
Yeah, that makes perfect sense, and I’ve run into similar issues with percentage-based metrics before. The unit picker is great for raw counts, but once you start mixing in calculated fields—especially ones that depend on weighted relationships—it just averages everything blindly and the logic golf hit behind the percentage gets lost. Your example highlights exactly why the results look off. Disabling the unit picker when advanced functions are used sounds like a reasonable safeguard, or at least there should be a warning so users know the aggregated percentages won’t be meaningful. Good call bringing this up.
Besides honing your SQL skills, you should also take some time to relax with Drift Boss. It's a high-speed game that helps you regain balance after stressful hours working with dry numbers.
Besides honing your SQL skills, you should also take some time to relax with Drift Boss. It's a high-speed game that helps you regain balance after stressful hours working with dry numbers.
Replies have been locked on this page!